I n s t r u m e n t s

a n d

M e t h o d s

Fig. 7. Relative frequency of shear frame measurements, which showed
smooth orrough fracture surfaces, depending on the pulling speed.

strain ratefor brittle fractures(Narita, 1983, Fukuzawaand
Narita,1993;McClung and Schaerer, 1993).
The shearstrength range and the strain rate range of our
thin weak layers or interfaces were, 250 to 4000 Pa and 10^-
2_{to 10}-1_s-1_{respectively.}
Shearing andfracturingthe weaklayers by a fast or slow
pull action (within 0.1 s and 2.5 s respectively) produces
definitely a brittle fracture at high strain rate, followed by
a catastrophic failure forthe given weak layerarea, simi-
larlyasduring theinitiationoflarger snowslabsindry
snow conditions. In the fast pull case the critical displace-
ment rate amounts- according to the above data -to2.5
mm/s,in the slow pull case to roughly 0.4 mm/s.
As we see from the Fig.6,the measured stress at frac-

tureorstrength ofeach sampledepends slightlyonthe
way we measure(fast orslow) and if the rupture surface
was smooth or rough (covered with small or large ripples).
Such ripples indicate an imperfect shearfracture,result-
ing in higher force and strength values.
In general"slow pull"measurementsyield better results,
because the control of the pulling action, e.g. the direction
alongtheslopeline,isbetterguaranteed duringa slow
pull. Thereforeas wesee from Fig. 7, slow pullsresult more
often in smooth shear surfaces, which improve the results.
It is important to note that a fast pull as well as a slow
pull aresufficient to reachthehigh strainraterange, which

Fig. 8. Differences of measured fracture stress due to the measuring
method (measuring with or without acceleration).

guaranteesa brittlefracture.Thereforesuch shearframe
measurements are representative, prognostic tools for slab
formation processesandvariousartificialavalancherelease
mechanisms. Accordingto resultsof Schweizeret al. (1995)
a skier exerts by skiing (weighting or jumping) peak shear
stressesinthesameorderofmagnitudeasourstrength
values and also in brittle manner thus the most important
prerequisites fora skier triggered slab are approachable.
Wehaveto mention that aup to nowhiddenerror source
forshear frame measurements has been localised so far as
under certainconditions several force-peaks are visible on
aforce-timediagramafterthefracture.Commonshear
strengthmeasurementsreveal onlythemaximumpeak,
which sometimes may not represent the fracture peak. In

such cases onlya parallel measured displacement curve
may indicate the right peak,i.e. when the fracture really
happened.This error sourcecontributes generallybetween
10 to 20%, but especiallywhen very weak layers are meas-
ured(shearstrength<500Pa), this factmay doublethe peak-
strength value (Fig. 8).
The shear strengthhas two basic components: cohesion
and friction.The cohesion ismainly related to the bond
strength of the snow grains and the friction to the weight
ofthesnowlayerabove.Aslongaswe have aconstant
velocityofdisplacementthefrictionforcestayssmaller
than thecohesion force, thus thefirst peakis identical with

weak layer

131